3.4 Microbial taxonomic composition and diversity measures
Amplicon sequencing of associated microbial communities on the Illumina MiSeq produced a total of 2,653,026 raw reads, and after quality filtering, dereplication and removal of non-bacterial sequences (Fig. S7, supporting information), a total of 6,249 different ASVs were identified. The dominant phyla detected across all samples were Proteobacteria (60.4%) (Fig. 7). In all samples from Brazil and some specimens from South Africa, Spain, and Australia, the phylum Firmicutes was also abundant (18.7%). In contrast, Bacteroidetes (7.8%) and Actinobacteria (5.5%) were observed only in some individuals across the localities. Noteworthy, specimens from Brazil exhibited a higher relative abundance of Firmicutes and Bacteroidetes than all other locations (Fig. 7). At the family level, Enterobacteriaceae  (27.4%), Anaplasmataceae  (4.4%) unclassified Clostridiales  (3.1%)  and Acetobacteraceae  (1.8%)  were predominant across all populations.
The individuals collected in South Africa and Brazil presented the lowest and highest number, respectively, of unique bacterial ASVs (581 vs. 2,526) (Fig. S8A-C, Supporting information). South African individuals shared a higher number of ASVs with the Palearctic and Australasian locations (73 and 98 ASVs, respectively) than the Neotropics (42 ASVs) (Fig. S8A-C, Supporting information). Individuals from South Africa and Brazil were collected from the same host plant, guava, but surprisingly the specimens collected from Brazil shared more ASVs with its neighbouring country Colombia (89 ASVs), despite being collected from different host plants, than with their putative native region in South Africa (52 ASVs) (Fig. S8B, Supporting information). This suggests that the bacterial composition remained similar to the native source in some localities, despite the medflies being collected from different host plants and geographically distant areas (see Fig. 1 and Supplementary Table 1).
Based on Shannon and Inverse Simpson diversity indexes, samples collected in South Africa appeared to have the lowest microbial diversity. At the same time, flies collected from different host plants and geographically distant sampling sites such as Israel, Spain and Australia showed the lowest disparity in diversity indexes, suggesting a similarity in microbial diversity. Conversely, the Neotropical region tended to have a more diverse microbiome, and samples from Brazil presented the highest diversity indexes (Fig. S9, Supporting information).